R, WIEBEAND V. L. GADDY
2300
a B-N-B pattern of linking, rather than of those containing B-B bonds; the B-N-B skeleton explains why the compound, when heated, gives good yields of B S N ~ H a~ , substance containing /N-B \ the B N ring; its rapid hydrolysis in acid \N-B/
solution to give five volumes of hydrogen suggests the existence of five B-H bonds. Finally, attentioil is called to a recent paper by S. H. Bauer,14 who, a t our suggestion, investigated the electron diffraction of the vapor of this compound, and concluded that the data obtained can be explained only by the existence of a B-K-B skeleton for the molecule. Acknowledgment.-We wish to express our thanks to the Research Corporation, for providing the liquid nitrogen used in a large part of this work.
Vol. 60
stance is hydrolyzed easily in acid solution, to give five volumes of hydrogen, two equivalents of boric acid, and one equivalent of ammonium ion. Its thermal decomposition (slow at room temperature) produces diborane and a solid of undetermined character. It reacts with an equal gas volume of trimethylamine, to form a stable white solid; on heating with an excess of that amine, this product yields borine trimethylammine. With ammonia, B ~ H T Nforms the stable solid B2H7N.NHs, which, on heating to ZOO', gives a good yield of triborine triamine (B3N3116). The ammonia addition product reacts with sodium in liquid ammonia to liberate one equivalent of hydrogen; the residue after thorough removal of the solvent corresponds to the formula NaNHy B2H7K. H H H
The structure B : N : B : H is proposed for the H H H
summary
compound, and although this specificpicture is not The new volatile compound B2H,N (b. p. 76.2'; considered definitely proved, it is shown to interm. p. -66.5') was prepared by the action of di- pret satisfactorily all of the physical and chemical borane upon its "&ammoniate." The new sub- properties of the substance. CHICAGO, ILL.
(14) S H. Bauer, Tars JOURNAL, 60, 524 (1938).
[CONTRIBUTION PROM
THE
RECEIVED APRIL11, 1938
BUREAUOF CHEMISTRY AND SOILS]
The Compressibilities of Hydrogen and of Four Mixtures of Hydrogen and Nitrogen at 0 , 2 5 , 50, 100,200 and.300"and to 1000 Atmospheres BY R. WIEBEAND V. I,. GADDY This is a continuation of work in this Laboratory on the compressibilities of hydrogen, nitrogen and their mixtures.l A summary of work previous to 1930 was given by Bartlett. Townend and Bhatt measured the isotherms of hydrogen at 0 and 25" t o 600 atm.2 Isothermal measure( 1 ) (a) Bartlett, THIS JOURNAL, 49, 687; 1955 (19271, (b) Bart lett, et al., ibid.. SO, 1275 (1928). (c) SP, 1363 (1930). Dr. Bartlett has kindly informed me that in Ref. l b the column of figures for the compressibility of a 3 : l mixture of hydrogen and nitrogen at 9A 85" was wrongly transcribed and shotild be as fdlow4. Press., . Old atm. figures Correct 1.3656 Also in the same table, for 300 1 1.3656 1.4027 atmospheres and Oo, the figures 50 1.3992 100 1.42g8 1.4364 have been transposed 200 1.5068 1.5170 300 1.5870 1.5975 Old Correct 400 1.6700 1.6799 figure figure 600 1.8412 1,8472 1 0264 1.2064 2 0130 2.0168 800 1000 2.1865 2,18135 (2) Townend and Rhatt Prof Roy SOG !London), 8 1 3 4 , i o ? (1981)
ments on hydrogen between 0 and 100" and up to 1000 a h . were made by Michels, et aL3 Michels and Gerver recalculated the compressibility data of Kohnstamm and Walstra on hydrogen a t 16.3 and 20' between 1000 and 2000 atm.4 Isotherms of nitrogen between 0 and 150° and at pressures from 20 to 80 atm. were determined by MicheIs, et aLS This latter work was continued up to 400 atm. by Otto, Michels and Wouters.$ A further extension to 3000 atm. including a calculation of the thermodynamic properties was made by Michels, et al.? W. Edwards Deming and Lola S. Deming have calculated the thermodynamic properties of hydrogen and nitrogens (3) Michels, Nijhnff and Gerver, A n n Phys , 18, 5G2 (1932, ( 4 ) Michels and Gerver, d i d , [5] 16, 745 (1933). ( 5 ) Michels, Wouters and De Boer, Physica, 1, 587 (1934) (6) Otto, Michels and Wouters, Physrk. Z., 86, 97 (1934) (7) Michels, Wouters and De Boer, Physica, 8, 585 597 (1936) (8) Deming and Shupe, Phys. R e v , 37, 638 (1931). thrd , 40, 848 (1932). netnina and Dernina, chid, 46, 109 (1934)
Oct., 1938
cOarpR12SSIsILITIES OF
HYDROGEN AND HYDROGEN-NITROGEN M m m s
Since in the previous investigations on hydrogen =me disagreement exists among the different ob-ers, we h ~ repeated e most of their measurements. A ~ a r a t u and s hedun.-Figure 1shows the low pressure part of the apparatus and the high presure pipet. Before each mu the connecting tube between pipet outlet valve M and stopcak G is open to the atmosphere through J. The 4-bulb buret A of 3000-c~.capacity is filled with KIand , J are closed, and gas a t same mercury. stopcocks € definite temperature and pressure is expanded from D through valve M. The whole volume is brought to atmos pheric pressure by means of the oil-manometer F. Stopcock G is now closed and the gas is confined to the &bulb buret A. The mercury levels are now adjusted to calihrationmarks betweenthe bulbs. Sinceusually twocombinations of bulbs are possible. the one giving a P m nearest to that of the atmosphere is selected. After opening H. manometer E is adjusted sa that the mercury surface in the right arm just touches the tip of pin L. This assures a constant volume betaem Hand L. The fmst two or three samples have to be discarded; after that the manometer will show the actual experimental variation, since the manometer volume is small compared with the gas volume in A. Any slight change in atmospheric pressure between closing of stopcock J and adjusting of oil-manometer F introduces only a negligible error since here again the volume of the connectingtube is small. Even a smaller volume is exposed to possible changes in mom temperature. B is a coil to bring the incoming gas rapidly to the temperature of C. The large thermostat C was kept at 25 * O.0lo while the temperature of the thermastat containing D was varied from 0 to 300'. Toluene and mercury filled regulators were used to COW the m g e between 0 and 200" while at 300" a thyratron coutrol was usea? We used the control only for an auxiliary immersion heater; the voltage a ~ o s sthe beating coils was that of the unregulated 110 a. c. voltagemains. Itwaspmible to maintain the temperature within a few one-hundredths of a degree with little trouble. The true temperatures WM measured with platinum mistance thermometers built according to spechications by Myers.w The thermometers were calibrated both at this Laboratory and at the Bureau of Standards with identical d t s . At 100 and criscO, or its equivalent. was used for liquid; while a t 3W0 a lead-tin mixture VB?I found satisfactory. our 200 and 1500 amwhere piston ([BBes pared at 100 atm. with the N a t i o h B A L of Standards gage," and it was found that only one of the Pistons had
2301
developed a change of one part in 10,oOO since August, 1935. The acceleration of gravity a t the laboratory is taken to be 980.091 cm./see.' The prwures are given in atmospheres, The thermal conductivity and an accurate combustion method" were used to analyze the gas m i x t m . The latter method was the more satis factory in OUT ease. The h i h PIvolume was measured at 25 and 100' and at 100 atm. in a slightly modhied steel pipet by dwlacing mercury by gas. The mercury was withdrawn. weighed and the volume dculated. The pressure coefficient was obtained from the work of Smith and Keyes.1: From several determinations the compressibility of hydrogen was determined a t 25 and 100". and 100 atmospbem. These values were used as reference points throughout this work. The average coefficients of linear expansion of our steel (C 0.49%, Cr 2.48%, V 0.25%. Mn 0.68%) were determined by Dr. Peter Hidnert of the National Bureau
Fig. 1. of Stan-. to whom our thanks are due. Table I givethe recults obtained.
TABLE I Temp. ran r.
02.
20-100
W200 20-300
Av-e
eaffidmt. of1iac.c expansion per "C. .?ample heated Sample treatcd r i f h t O 300' hydrogen at 300"
12.4 X 10" 12.9 x 1013.6 X 10-
11.4 X 1012.2 x IO" 12.9 x 10-
---
(9) Zabel md Hmem,h. Sci. IWrYmnIs, 8, 28 (1934). (IO) Myen. Bur. SLmdmds J . fisewch, 9, 807 (198% ( I t ) M w c n and J ~ u Piw.. . 8.10(11(1881).
The results of Table I show a decided influence of hydrogen on steel. The sample used in determining the linear coefficient of expansion had been exposed to extreme conditions. In actual pradice the thickness of the pipet walls (2.5 an.) and the short time exposure at 300" and loo0 (12) Shepherd. W..0,121 (1981). (13) Smith and Kern. Roc. A n . I c d . AIL? S&. 69, 818 (1984).
R. WIEBEAND V. L. GADDY
8302
Vol. 60
TABLEI1 EXPERIMENTAL pu V u m s Hydrogen
87.44 : 12.56 Hr N2 mixture
75.56 : 24.44 Ht:Nr mixture
51.74 : 48 26
F h Nr mixture
26.12 .73.88 A2 Nz mixture
Nitrogen
1 0062 1 0165 1 0416 1 1081 1 2837 1 4802 1 6820 1 8824
0 9999 1 0038 1 0191 1 0789 1 2728 1 5023 1 7382 1 9727
0.9910b .9853b ,9857" 1.0363" 1.2566" 1.5251" 1.7984" 2,Ofi76"
1 1015
1 U94-l
1 1130 1 1425 1 2127 1 3878 1 68lfi 1 7822
1 1020 1 123f) 1 1894
0" 1.0152 1 0311 1 U641 1 1327 1 2761" 1 4221U 1.5668" 1 708ti"
1 013b 1 0293 1 0613 1 1311 1 2829 1 4411 1 5994 1 76Ki
1 1072 1 123.5
I 1063 1 1224 I 1583 1.2272 1 3794 1 5372 1 6948 1 XBO8
1.0265 1.0572 1.1273 1.2863 1 4558 1 6282 1 7989 2*3O
1 1572
I 2271 I 37rl.7 1 51,5h i t7.595 1 Xl?l.
